Retantate
O-ring
Thesis III Chapter IV
59
4.3.3. Modifications
Jacketing of PV unit to keep temperature constant
Due to heat lost and effect of it on selectivity, the unit was jacketed at workshop of institute and later it was connected to the thermostat to keep the temperature constant at 40°C (Figure 4.17.).
Thermostated water inlet
Membrane Outlet
Figure 4.17.: Representation of heat system of the experimental equipment.
Inlet modification of PV unit
Since the inlet and outlet of the PV unite was close to each other, some modifications were done to increase the flux. Firstly the inlet was moved from centre of PV unit to the left side and the previous inlet was sealed (Figure 14.18.). However the result was still not satisfying.
Inlet Inlet
After modification
Figure 4.18.: The first modification of the membrane unit
Secondly the inlet was introduced to PV unit tangentiallyand previous inlet sealed (Figure 4.19.).
Thesis III Chapter IV
Inlet
Inlet
Figure 4.19.: The second modification of the membrane unit
In this way the effective membrane surface area was enlarged considerably, as it can be seen in Figure 4.20.
Before any modification I. Modification II. Modification
Effective membrane surface areas
Figure 4.20.: The effect of the modifications on the membrane test cell
It can be considered in Figure 4.20., that after the final modification the effective area of the membrane was ten time larger than before any modification. Thus almost all the area of the membrane can be used in the further applications.
Thesis III Chapter IV
61
4.3.4. Characterisation of the pervaporation membrane
Applying pervaporation for a given separation, the most important thing is to find a proper membrane. Then it has to be characterised experimentally. The two main parameters for characterisation are the flux and the selectivity of the membrane.
Pervaporation can be used for separation of both products in the esterification reactions. However, in this section recovery of only the ester component from the reaction mixture having no solvent (solvent free system) is aimed. The ester compound, ethyl acetate – having hydrophobic feature – can be separated from the reaction mixture by hydrophobic (organophilic) membranes. Unfortunately only one type of orghanophilic membrane available commercially was suggested to apply for our purposes: the PV 1060 type membrane from GFT (now Sulzer, Germany). So our experiments were carried out with this membrane, using model solutions (ethyl acetate – ethanol, containing no water, acetic acid or enzyme)
Procedure of the experiments. Prior to the characterisation of the PV membrane, firstly its compatibility should be checked. Therefore it was immersed into the particular reaction mixture, containing ethyl acetate and ethanol. It was found that the compounds present did not damage the membrane, they were not harmful. Moreover, the gasket material has to be stable against the components of the feed mixture, as well (which was also checked) and must be free from apparent damages.
Then the membrane was cut into a disc form suitable for the test cell (diameter 158 mm). The membrane must not handle with naked hands, it is necessary to use gloves when holding the membrane. Folding, dirt, bending and punching should be avoided to keep the membrane unharmed, intact, defect-free.
The PV membranes have to be conditioned before using, this was kept in the reaction mixture to be separated for a couple of hours. After the permeate condensers were
Thesis III Chapter IV
installed under the test cell, the disc from PV membrane was adjusted carefully onto the top of the lower part. The o-ring must fit tightly into the test cell. Then the cell was closed by fixing and fastening the bolts and nuts.
As a next step the cell was connected to the vacuum pump and the vacuum of the system was checked, to study if any leakage occurs. Similarly the thermostat operation was checked and the jacketed parts of the test cell were filled with the heat carrier liquid (water). Between the permeate side of the unit and the vacuum pump, there are three dry ice acetone traps to collect the vapours that permeate through the membrane. Those cold traps are joint parallel to condense more permeate if necessary. During the experiment those traps were collected and heated it up to the room temperature to measure the weight.
The vacuum pump was inline with columns to avoid the vapour getting into pump. To measure the pressure manometer was used. The scheme of the system is presented in Figure 4.21.
Figure 4.21.: Scheme of the system
membrane thermostate
Thesis III Chapter IV
63
During the tests 90 w/w% ethanol and 10 w/w% ethyl acetate mixture (200 ml) was circulated by a peristaltic pump in the primary (upper) side of the cell at 40 °C temperature and the vacuum used was 8 kPa. The effective membrane surface area was 178 cm2. Permeate samples were collected in the traps and their weights and compositions were determined for the calculations of selectivity and flux, respectively. The composition of the permeates as well as the retentates was measured by a GC method. The tests were repeated five times, and the average was calculated.
Experimental results. The results of one representative experiment (as an example) are summarised in Table 4.6. The selectivity and the flux data were calculated from the composition (by GC analysis) and weight of the permeate samples, respectively.
Table 4.6.: The result of the GFT PV 1060 membrane test experiment
Time
Thesis III Chapter IV
0 1 2 3 4 5
selectivity
0 1 2 3 4 5 6
sample
The selectivity and the flux as they were defined in chapter 1.6., page 20, obtained are shown in Figures 4.22. and 4.23. It can be seen that ethyl acetate was concentrated in the permeate (compared to the feed) so the PV membrane is suitable to enrich the mixture in ethyl acetate. It can be seen that the selectivity and flux were changed as a function of the measurement time. Its reason is the changing in composition of the feed, thus the driving force has changed, as well. Therefore only the first data were used in our further calculations.
Fıgure 4.22.: The selectivity of the pervaporation membrane
Thesis III Chapter IV
65
As a result of our investigations it was found, that the average selectivity of the PV 1060 membrane was 5.1, while its flux can be characterised as 1.6 kg/hm2 (as an average).
Time [min]
0 20 40 60 80
Flux [kg h-1 m-2 ]
0.0 0.5 1.0 1.5 2.0
Figure 4.23.: The flux of the pervaporation membrane
Thesis IV Chapter IV
4.4. Enzymatic esterification in integrated system
Statement: